News Stay informed about the latest enterprise technology news and product updates.

Q&A: Nortel's data networking chief discusses the future

Recently, published a story about the future of the Nortel enterprise data business in the wake of the company's bankruptcy. The story characterized Nortel's data portfolio as an old product line. The story also said that the many acquisition rumors surrounding the company focus on the notion that any buyer of Nortel's enterprise business would be more interested in its customers than its technology. John McHugh, Nortel's vice president for data solutions, requested an opportunity to defend his company's data networking business and to discuss its future. The result is this Q&A.

Do you hear from your customers any concerns about whether they're going to lose their sales reps, their sales engineers, their support guys?
What I'm hearing is customers telling me how much they respect [our people] and how valuable and unique the folks that call on them from a sales, technical support and services standpoint are -- how much they are respected and valued. So that is something we have to consider very carefully in all of this. Economically across the board, everyone is having to make very difficult decisions about how to get their businesses right-sized for the economy. And we're going through the same situation. If there's any benefit that Nortel has, it's that we frankly started earlier than everyone else, and I think we're a little further along in the process. Being new to Nortel, I've spent much of the last six months talking to customers. And I've been very impressed with how tight the relationships are between our customers and the folks who are selling and are directly interfacing with our customers.

I have a thick skin. This business unit frankly has far more credibility and depth than a lot of people are giving it credit for. There is another perspective -- of a healthy business that's going through something where the stewardship probably needs to be changed. The grand experiment of Nortel acquiring Bay Networks may not have been the best strategy in the world. And I think there is a lot that can be learned by why there aren't any traditional voice companies that pursue data networking that can point to success. It's a model that largely hasn't worked. And data networking companies persist in being very focused, value-added strategic providers to their customer base. We see ourselves as getting back on that course, even in a world where the new bright shiny object is data center networking solutions. My perspective is, if you are a networking infrastructure player and you run too hard in that direction in the reinvention and re-architecture of the data center, I think you are going to overcorrect and become a niche player. At the end of the day, customers don't buy networks just to support unified communications or just to build the data center. They buy a network because it's got to support all of an enterprise's applications.

Your company objected to the characterization of Nortel's data networking portfolio. Could you elaborate on how you disagree with the story?
I guess more than anything, the article tends to convey a Nortel data networking business which was not the product of much investment, not really relevant or very impactful.

It really goes against what most of the market share information will tell you. This business in North America is still one of the top three or four leading vendors for networking infrastructure in terms of switching and infrastructure deployments. You get the impression that … no one is very interested in where this business is going to end up. I think it's quite to the contrary. I think there are a lot of customers out there who have a significant amount of Nortel infrastructure, and we have been investing pretty heavily. In fact, if you look at 2008, Nortel's investment in enterprise networking infrastructure was second only to Cisco. We've refreshed our entire line in the last 12 months.

Nortel was second to Cisco in terms of research and development last year?
Correct. Last year, we invested something approaching $200 million in the business that was going into core routing platforms, virtualization architecture, security, routing, wide area, wireless, as well as our switching platforms, both modular and stackable. It was 10% to 15% shy of $200 million.

If you look over the industry -- at a lot of the talk and a lot of analyst commentary -- obviously there's Cisco, and everyone knows Cisco's place. They step back and say, "Who is relevant, and who is a comprehensive player that can build out an end-to-end solution? Right now, the only player in the industry that can build out an end-to-end solution to the market is Nortel. Juniper has made some progress. They brought their first-generation switches to market last year. They still don't have anything in advanced virtualization [or] anything in core-routing switches that's very comprehensive or integrates widely into end-to-end architectures. HP [ProCurve] is just an access layer player. Even Foundry [Brocade], which comes the closest, doesn't have much of a remote office solution, or wide-area solution, or VPN strategy in its architecture. Even its wireless is mostly outsourced.

More on the fall of Nortel
Nortel bankruptcy: Timeline of a hard fall

Rivals of bankrupt Nortel to poach customers aggressively

At VoiceCon, Nortel customers uneasy about vendor's future
You mentioned that you are still in the top three or four of networking markets today. What about market trends? Where is your market share heading right now?
It's certainly kind of hard to pick that up in 2009 at this point. Nortel's market share going back three years, which is one of the reasons I'm leading this business unit now, has been mostly flat. The process we're going through right now is to cure the [problems from the fact] that Nortel really for the last four or five years did not position this business as a standalone, competitive data networking business that was known for its differentiated value and its differentiated capabilities. It's resulted in this business unit being very effective selling into its base but, frankly, not growing. And when the market was showing growth, this business unit was not participating and not taking market share as it should have been. You can look at Dell'Oro's numbers or IDC's numbers, and Q1 was a bloodbath across the industry. I think we had the added impact from the fact that we filed for Chapter 11 during that time. But our numbers were not dramatically different from the overall downturn in the market. Earlier you said the Nortel data network portfolio has gone through a total refresh in the last 12 months. Could you elaborate? Some analysts have characterized the Nortel ERS 8600 as a dated platform.
That's the high end of modular core solutions today. While it's true the 8600 was originally introduced in its initial variant in 2000, not unlike the [Cisco] Catalyst 6500, both have been refreshed in several substantial ways since then. We did do a complete architecture and performance refresh in 2008. That platform is basically completely refreshed, and in its current form I believe it's got a relevant performance profile to be able to support gigabit assignation roles and 10 gigabit core roles for enterprises for the next five to 10 years. It's a very well-structured architecture. Compared to the Catalyst 6500, it's got better value and better performance.

We also refreshed our stackable line. One of Nortel's unique contributions to industry is to have a genuine, high-availability, hot-swappable stacking architecture for wiring closet deployment. In January, we released a new top-of-line product, the ERS 5600. We now extend that stackable capability to a terabit. So you can literally build a terabit stack of these stackable 1 gigabit devices with full resiliency, failover, hot swappability, and unified management around those. There's no one in the industry that comes close to that architecture and capability. Even the lightweight chassis that HP sells does not have that kind of backplane performance to be able to build out a data center top-of-rack solution or wiring closet. So it's not just new stuff. It's highly differentiated.

You mentioned Cisco's Catalyst 6500, but how would you say you stack up to Cisco's Nexus family?
The VSP 9000 is comparably positioned to the Nexus. And that's important in a couple of ways. The Nexus is not replacing the 6500, and the VSP 9000 doesn't attempt to replace the 8600. Nor is it seen as necessary for the 8600. The VSP 9000 is really a different role. It's a 10 gigabit aggregator, first of all. If your objective is to put lots and lots of gigabit ports into a product, you wouldn't want to do that on the Nexus, and you wouldn't want to do it on the VSP 9000 either. The VSP 9000 is a virtualizing engine as well. So as we look at more functionality in the data center -- in high-end core applications and even in light-duty metro applications, where you have more virtualized network services actually residing in the network -- this is the architecture that's going to support that. The thing that some analysts are confused about [is that] they see the Nexus as, "Oh, it must be … the replacement of the 6500. Nexus must be the next thing." It's really not. What the VSP 9000 is doing is even beyond what Cisco is doing. It is in fact a code extension to the 8600. The functionality that is on the 8600 should also immediately show up on the VSP 9000. Customers will use it as a performance upgrade, but they will also be able to coexist inside the same architectures for such advanced technologies as SMLT [Split Multi-Link Trunking] and PLSB [Provider Link State Bridging]. It's more like a big brother of the 8600. And when will the VSP 9000 be coming?
That product will be in trials at the end of this year. Given the current situation, where Nortel has made it clear that all of its assets are for sale -- if that happens, it could happen before the VSP 9000 comes out. Will that switch really reach the market if the enterprise business is sold? Is the product going to ship in 2010 regardless of who owns the division?
When I was leading the ProCurve business at HP, there were five years where it was the worst-kept secret in the industry that HP was trying to divest of that business, and yet that transaction never occurred even though it was often speculated about. What I learned and what really is persistent in these situations where there is ambiguity around a business model, is that the persistent things are the assets of the business. We have a workforce and a strategy which is working very well. We're seeing very strong performance in the second quarter of this year because we're not wasting a lot of time speculating about what's going to happen and changing our strategies based on what may or may not happen. We know at the end of the day that if you have competitive products and you're selling a lot of them, if they're high quality, if you've got satisfied and loyal customers, if you have things that are assets, they're going to persist through the transaction. The one thing [a buyer of the division] will do is protect the assets. The VSP 9000 represents a nearly $200 million investment. It's within six months of being in trials. If you've got an installed base which is arguably the second largest in North America, all running on an 8600, this represents a and line extension…

We showed this at Interop, and I had customers coming in day in and day out. They were literally spending 30 minutes thinking through the timing of when they can deploy it and how they can deploy it. There is very fertile ground for this architecture out there. I've got to think anyone looking at the Nortel Enterprise Solutions organization as a potential acquisition has got to see the asset potential of this base networks business, which has been persistent, which has had various levels of competitiveness and is really well poised for the next five to 10 years [to get] back into the mode of taking market share and growing.

So you're seeing strong sales in the second quarter? What are people buying?
It's fairly well split. The data networking business of Nortel, typically your stackable solutions, depending on what supplier you look at, represents about 80% of units and revenues sold. The module solutions, or the core level of products, represent the bulk of the remainder, somewhere between 15% and 25%. Nortel has always been unique because we have a very integrated approach to resiliency, and we're always building on network architectures. So our core products have represented a higher proportion of our overall revenue than other suppliers in the industry. It's been interesting to see how Chapter 11 initially impacted our non-core products much more directly. There was an intake of breath as people challenged what their tactical decisions were but actually didn't challenge their strategic buying as much. That has actually rebalanced this quarter, and we're seeing stackables coming back. It's interesting that you're seeing a strong quarter, given the uncertainty with the company. How are you winning those sales?
The uncertainty is a lot more in terms of how people are perceiving the company rather than the relationships customers have with the company. So what customers are seeing day in and day out is the same Nortel representatives that have been selling to them in many cases going back to the Bay Networking days. They're seeing the same technical support folks. They're seeing the same very comprehensive support capability. There's only one company in the industry that I think can tout the same level of reach and touch that Nortel has [Cisco]. Customers keep reading about disruption and discontinuity and ambiguity and a questionable future, and yet what they see is the reality around how deep this asset is and how competent the business capability is. I'll be honest with you. We're selling into our installed base as we have been for the last five years. Our comfort zone, our sweet spot and our effectiveness has been maintaining and expanding those ongoing relationships. But we have new customers as well.

Our biggest issue is getting through this noise and fog that's been created by this complex [bankruptcy] process and reminding customers that the underlying asset is a healthy one. It's got lots of continuity, and this is more curing some of our problems than it is the cause of some of our issues.

Do you have a sense how any transaction might affect operations in the division? You must be saying something to your customers, right?
What I'll tell you is I was brought into Nortel because I have a very deep background and experience in carve-outs and planning for operational continuity. I went through the Compaq-HP merger and am very familiar with the process by which you take very complex organizations and maintain coherence of operations through a disruptive period. It's a lot around the diligence and sweating the details and understanding how to maintain operations while planning for a transition. It's a challenge, but it's also not rocket science.

Dig Deeper on Data Center Networking

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.